ESP_DMX on ESP32-S2 about esp_dmx

[someweisguy]

The latest commit on release/v3.1 compiles and awaits RDM data without issue. I haven't tried testing any RDM receiving yet.

[fedechi]

ok, it was my mistake really sorry. I inherited from the previous examples the initialization in the form
ESP_ERROR_CHECK(dmx_driver_install(dmx_num, &config, DMX_INTR_FLAGS_DEFAULT));
but having changed the return of the function obviously nothing worked anymore!
Now dmx works, the code run but unfortunately I still have problems on RDM discovery with my test setup.
I can't discover with ENTTEC RDM controller, the therminal says
W (20592) dmx: PID 0x0001 did not send a response
With the oscilloscope I can observe that after idle state of DMX line, when responder receive the first RDM_PID_DISC_UNIQUE_BRANCH
to which he must answer, the responder switch to TX mode the dmx line quite early (<176us) but writes nothing. After 25ms the line return to RX state.
Also through RDM packet sniffer i can see an error on checksum for ACK response to DISC_MUTE command.
thanks
F.

[someweisguy]

Interesting. Thank you for the detailed information!

I will need to do some more in-depth research into this problem. Unfortunately, I am in the middle of moving across the country so I cannot access my testing equipment at the moment.

This library uses a timer to keep track of when the latest DMX/RDM data was received. When the ESP32 receives an RDM request, it will attempt to send a response. It sounds like you are already aware that RDM requires that responses to packets be sent after about 176 microseconds and within about 3 milliseconds. If more than 3ms have elapsed when the library is attempting to send an RDM response packet, this library prints a warning message. This is why you are seeing W (20592) dmx: PID 0x0001 did not send a response. It seems like there may be an error in the way this library configures the timer on ESP32-S2.

How many RDM devices do you have connected on your DMX network? It appears you may have at least 3 different devices: the ENTTEC RDM controller, the ESP32-S2, and a different device manufactured by a company called ERAL srl. This company appears to also be called Linea Light Group. The device manufactured by ERAL srl/Linea Light Group is the device that is returning a bad checksum. The source ID of the ESP32-S2 should begin with 05E0 unless you changed it using the ESP-IDF sdkconfig.

[fedechi]

no problem, have a good travel, we will go depth on all the issues in september. In the test there is only one device with esp_rdm lib and yes, i have modified UID from sdkconfig. Happy holidays.

[someweisguy]

Issue #86 has been resolved, so I am hoping that the fixes in that issue have carried over to this issue. I will do some testing soon now that I have access to my ESP32-S2 again!

[someweisguy]

This appears to be resolved on both ESP-IDF and Arduino. Let me know if you run into any more issues!